Spell Blind Spot

## Summary

**Systematic blind spot detection.** Takes a topic — a plan, decision, technology, or research question — and probes it from five angles to surface hidden assumptions, unexamined failure modes, missing perspectives, and cross-domain insights. Produces a structured analysis with honest coverage assessment.

**Arguments:** `$ARGUMENTS` (required) - Topic, plan, decision, or question to probe for blind spots

**Output:** Structured blind spot analysis with findings, pattern map, limitations, and specific next steps output directly to the conversation (Write is available if the user requests the output be saved to a file)

---

## Skill References

Before proceeding, load the relevant skill documents for reference:

Read the following files from the repository root:
- `docs/skills/assumption-surfacing.md`
- `docs/skills/unknown-unknowns.md`

---

## Process

### Phase 1: Frame

#### Step 1: Parse Arguments

**If `$ARGUMENTS` is empty:**
Ask the user: "What topic, plan, or decision should I probe for blind spots?"

**Otherwise:**
- Extract the topic, plan, decision, or question from `$ARGUMENTS`
- Continue to Step 2

#### Step 2: Map Known Territory

Establish a baseline of what is already known before probing for what's missing.

**Web research (if available):**
Search the web for current knowledge, known challenges, and documented lessons learned for the topic.

**Codebase analysis (if relevant):**
- Check if the topic relates to the current codebase — look for relevant code, config, docs, ADRs, or prior decisions
- Identify what the project already handles and what it doesn't

**Graceful degradation:** If web search is unavailable, rely on built-in knowledge and codebase analysis. Note this honestly in the Coverage field.

**Output the baseline:**